Thrombosis after covid-19 vaccination
BMJ 2021; 373 doi: https://doi.org/10.1136/bmj.n958 (Published 14 April 2021) Cite this as: BMJ 2021;373:n958Read our latest coverage of the coronavirus outbreak
Linked Opinion
Maintaining trust in the covid-19 vaccine programme will be essential in the months ahead
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Dear Editor
Both the vaccines and CoVID19 induce a very similar atypical coagulopathy (1). As the spike protein is the only feature in common, it seems likely that the pathogenic mechanism for both the rare vaccine side effects and infection involves this viral component. An unusual feature of the SARS-CoV2 Spike is that the protein is heavily glycosylated in both virus and vaccines and the glycosyl moieties carried are variably dependent on the host cell in which the protein is expressed (2). The importance of the innate immunity in SARS-CoV2 infection is generally recognised and, apart from the interferon response also includes a number of other plasma proteins, most notably Complement (3) which has recently been implicated in the activation of the coagulation cascade (4).
It is thus interesting to note that one of the three pathways of the activation of complement is via the alternative pathway, is triggered by the detection of aberrant glycosylation patterns found in both host cells and pathogens (4), Thus a simple pathophysiological mechanism for the coagulopathy observed in both CoVID19 and the rare vaccine reactions is that the aberrant activation of the alternative pathway of complement, triggered by hyper-glycosylated Spike protein, and this indeed has been directly demonstrated (5).
Moreover, it is intriguing to note that in the longitudinal analysis of immune response markers in CoVID19, the activation of Complement was seen in mild SARS-CoV2 infections, before the appearance of the immune response makers that correlate with more severe disease (3)
1. Cines Douglas B., Bussel James B.. (2021) SARS-CoV-2 Vaccine–Induced Immune Thrombotic Thrombocytopenia. N Engl J Med DOI: 10.1056/NEJMe2106315.
2. Brun J et.al. Analysis of SARS-CoV-2 spike glycosylation reveals shedding of a vaccine candidate bioRxiv 2020.11.16.384594; doi: https://doi.org/10.1101/2020.11.16.384594
3.Bergamaschi L et.al. Delayed bystander CD8 T cell activation, early immune pathology and persistent dysregulation characterise severe COVID-19. medRxiv 2021.01.11.20248765; doi: https://doi.org/10.1101/2021.01.11.20248765
4. Kenawy HI, Boral I Bevington A Complement-Coagulation Cross-Talk: A Potential Mediator of the Physiological Activation of Complement by Low pH. Frontiers in Immunology, 06 May 2015, 6:215. doi: 10.3389/fimmu.2015.00215
5. Yu J, Yuan X, Chen H, Chaturvedi S, Braunstein EM, Brodsky RA. Direct activation of the alternative complement pathway by SARS-CoV-2 spike proteins is blocked by factor D inhibition. Blood. 2020 Oct 29;136(18):2080-2089. doi: 10.1182/blood.2020008248.
Competing interests: No competing interests
Dear Editor,
Comparisons have been drawn between the AstraZeneca vaccine, and the combined oral contraceptive pill with regards to the likelihood of a thrombotic event occurring. Thrombosis associated with the combined oral contraceptive pill is estimated to affect around 1 in 1000 women, which is significantly higher than that of the AstraZeneca Vaccine – 1 in 25000 (Lantos, 2021). It should be noted that no widespread criticism of the pill has emerged since the reports of thrombosis in the AstraZeneca vaccine, yet many have been hesitant and even opposed to the administration of the vaccine, due to the reports of these rare thrombotic events.
Perhaps then, the nationwide hesitancy over the AstraZeneca vaccine is rather as a result of bias in media coverage. This bias may stem from the unique situation of the AstraZeneca vaccine, being one of the first to pass through clinical trials at such speed during an unprecedented pandemic, and the speed of rollout has drawn particular attention to rare yet severe side effects that might not have been covered by the media during other times. Increased focus on these rare thrombotic events should be put into better context.
The risk of central venous thrombosis is significantly higher after covid-19 than after receiving the AstraZeneca vaccine, and the risk of cerebral blood clots from the disease is 10 times that of the vaccine (Torjesen, 2021). The vaccine hesitancy has been highest in younger groups, with the spread of misinformation over social media being a significant factor (Mehta, 2021). The author’s summary that the members of the public should accept the vaccine, as the benefits seem to outweigh the risks, echoes the message that has been promoted by the NHS. Perhaps then a more balanced coverage of the statistics, and the benefits of the vaccine would reduce the speculation and resistance against the AstraZeneca Vaccine. Public health officials may be best placed to interact with the media to help deliver this to the general public.
Lantos, E., 2021. Covid: Women should 'keep taking the pill' amid jab clot fear. [online] BBC News. Available at:
Torjesen, I., 2021. Covid-19: Risk of cerebral blood clots from disease is 10 times that from vaccination, study finds. BMJ, p.n1005.
Mehta, A., 2021. COVID-19 vaccine hesitancy among black people in UK halves, new data suggests. [online] Sky News. Available at:
Competing interests: No competing interests
Dear Editor
Is tozinameran (the Pfizer/BioNTech vaccine) safer than ChAdOx1 (the Oxford-AstraZeneca vaccine)? The Oxford-AstraZeneca vaccine may cause an extremely rare and very distinctive syndrome of cerebral venous sinus thrombosis, thrombocytopenia, and antibodies against platelet factor 4, as Professor Hunter suggests, but we need to know the aggregate risks of harm from the vaccines, not this specific risk. We are unlikely to discover the response from spontaneous adverse drug reaction reports, which require that the reporter suspects an association between some harm and a vaccine, and then reports it. The number of adverse drug reaction reports concerning the Oxford-AstraZeneca vaccine grew by over 10% in the week to 5 April 2021, probably as a sequel to media interest. Even so, the MHRA data show that after approximately 15.4 million doses of Pfizer/BioNTech vaccine there have been 314 reports of fatal reactions (about 20 per million), while for the Oxford-AstraZeneca vaccine, after 21.6 million doses, 521 fatal reactions (about 24 per million) have been reported.
Until carefully conducted cohort and case-control studies show otherwise, it would be wrong to conclude that one vaccine is safer than another. Compared with most medicines, where a reaction that occurs in fewer than 1 in 10,000 is classed as ‘very rare,’covid-19 vaccines appear to be extremely safe. In the meantime, it will be best to report any suspicion of a serious adverse reaction to any covid-19 vaccine.
Competing interests: No competing interests
Dear Editor,
In a recent editorial about thrombosis following inoculation of AstraZeneca Covid-19 Vaccine [1], Prof. Hunter unfortunately misquotes the results of an observational study by Public Health England (PHE) available as a non-peer-reviewed preprint [2], stating that a single dose of the AstraZeneca vaccine was shown in the study to be effective at preventing death. The study authors explain instead that “there was insufficient follow-up to assess the effect of [the AstraZeneca vaccine] on mortality due to the later rollout of this vaccine [compared to the Pfizer vaccine rollout]”.
The editorial concludes with the statement that “it remains the case that the benefits of the AstraZeneca vaccine far outweigh the risks for all adult age groups, especially when the potential for long covid is considered”. However, in an earlier paragraph, Prof. Hunter writes that “the exception is the 20-29 year age group, for which the risk-benefit equation is more finely balanced when community transmission is low.” A comparison is then made between the risk of blood clots from vaccination and that from taking the contraceptive pill. The comparison does not add to Prof. Hunter’s argument that also for the younger adults the harms-benefits equation is favourable when taking the AstraZeneca vaccine. The two cases’ respective unwanted outcomes, that is, death from covid-19 and unwanted pregnancy, have extremely different absolute probabilities of occurring. Hence, the respective possible benefits from taking preventative measures against them are also extremely different.
A healthy woman in the 20-29 age group who wants to avoid pregnancy while fully enjoying a sexual relationship with her partner, would have a maximally high 1 in 1 risk to get pregnant if she does not take the contraceptive pill or other birth control measures (provided obviously that the couple is reproductively healthy). The same woman declining the offer of the AstraZeneca vaccine will have instead, according to a January 2021 PHE own report [3], an extremely small risk of catching and dying from covid-19 of less than 2 in 100,000. The healthy woman would in fact have a lower risk than this, as most 20-29 year old individuals dying from covid-19 have relevant comorbidities. Moreover, that woman can further reduce the risks associated with a covid-19 pandemic by exercising personal control on her behaviour in relation to infection prevention and control measures, even if she declines the vaccine. Conversely, once she gets vaccinated, she loses control on the known and unknown harms from the inoculation. Her body, her choice.
Most importantly, the European Medicines Agency now states in the package leaflet side effects section of the AstraZeneca vaccine (now relabeled Vaxzevria in Europe): “Very Rare (may affect up to 1 in 10,000 people) – blood clots often in unusual locations (e.g. brain, bowel, liver, spleen) in combination with low level of blood platelets” [4]. Hence, this risk is now given as 1 in 10,000 in Europe, contrasting with the more reassuring 1 in 250,000 in the UK. I have briefly provided in a previous rapid response the reason for the discrepancy possibly being partly due on the inefficiencies of the UK pharmacovigilance infrastructure [5].
Incidentally, it is interesting to note that news pieces often present the actual extremely small figures of risk of harms from the AstraZeneca vaccination, but rarely provide the actual extremely small figures of probability of benefits from vaccination – which is very different from the well advertised AstraZeneca vaccine efficacy figure – in certain population groups. Only with both figures a harms-benefits equation can be estimated. Instead, what is often presented are unhelpful comparisons with the risk of harms from common conditions or treatments, such as the contraceptive pill, unrelated to covid-19.
Finally, Prof. Hunter writes that “without the AstraZeneca vaccine it is doubtful that hospital admissions in the UK would have fallen as much as they have, delaying release from lockdown.” However, from the Office for National Statistics figures [6], the fall in deaths and hospital admissions that started just before spring 2021 broadly mirrors the fall we saw back in spring 2020, which occurred without the administration of vaccine to millions in a world population that presumably had overall low immunity to covid-19. Hence, it could be equally reasonable to give the merit of recent falling numbers to lockdown measures, which were in place on both occasions, or to a degree of herd immunity reached through previous infections since the pandemic outbreak or through natural immunity [7]. Or perhaps covid-19 has a strong component of seasonality, which could be related to several factors [8], possibly explaining at least part of the fluctuations in both years.
The uncertainty about the real cause of declining deaths and hospital admissions stems from the observational nature of the available preliminary analyses. The truth is probably that a combination of all the above factors is at play. However, with the current evidence, assigning to the AstraZeneca vaccine the bulk of the merit for allowing the easing of lockdown feels like mere speculation.
References
[1] P.R. Hunter, Thrombosis after covid-19 vaccination., BMJ. 373 (2021) n958. https://doi.org/10.1136/bmj.n958.
[2] J.L. Bernal, N. Andrews, C. Gower, J. Stowe, C. Robertson, E. Tessier, R. Simmons, S. Cottrell, R. Roberts, M. O’Doherty, K. Brown, C. Cameron, D. Stockton, J. McMenamin, M. Ramsay, Early effectiveness of COVID-19 vaccination with BNT162b2 mRNA vaccine and ChAdOx1 adenovirus vector vaccine on symptomatic disease, hospitalisations and mortality in older adults in England, MedRxiv. (2021) 2021.03.01.21252652. https://doi.org/10.1101/2021.03.01.21252652.
[3] Public Health England, COVID-19 confirmed deaths in England (to 31 January 2021): report, (2001). https://www.gov.uk/government/publications/covid-19-reported-sars-cov-2-....
[4] European Medicines Agency, Annex I - Summary of product characteristics (Vaxzevria), (2021). https://www.ema.europa.eu/en/documents/product-information/vaxzevria-pre....
[5] M.T. Suadoni, Pharmacovigilance needs improvement (Rapid response to https://doi.org/10.1136/bmj.n931 ), BMJ. (2021). https://www.bmj.com/content/373/bmj.n931/rr-16.
[6] Office For National Statistics, Deaths registered weekly in England and Wales, provisional: week ending 2 April 2021, (2021). https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarri....
[7] British Society for Immunology, Immunity & COVID-19, (2021). https://www.immunology.org/sites/default/files/BSI_Briefing_Note_2021_im....
[8] K. O’reilly, J. Edmunds, A. Bennet, J. Reid, P. Horby, C. Noakes, Seasonality and its impact on COVID-19, Sage 63. (2020) 1–10.
Competing interests: No competing interests
Dear Editor
Many women across the world after receiving CoViD vaccines are complaining of irregularities in their menstrual bleeding; some experiencing heavy menstrual bleeding (menorrhagia), some bleeding before their periods were due or bleeding frequently (metrorrhagia/polymenorrhea), whereas some are complaining of postmenopausal bleeding.
As of 5th April 2021, there have been ~958 cases of post-vaccination menstrual irregularities, including vaginal haemorrhages, that were recorded in MHRA’s adverse event reports. There were twice more cases of menstrual irregularities with CoViD Vaccine AstraZeneca than Pfizer (643 vs 315 respectively) [1]. It is anticipated that the actual numbers of cases are much higher than the numbers recorded in the pharmacovigilance systems as many women in different cultural context may have felt uncomfortable to talk about it, may not have thought that it was vaccine-related, or may have not been encouraged by their clinicians to make an official report into the adverse events reporting system.
There have been recent reports of haemorrhage, blood clots and thrombocytopenia following administration of CoViD-19 vaccines that have raised concerns over the safety of genetic vaccines for people with pre-existing coagulation disorders or those on certain medications. Regulatory bodies have also issued warnings to the patients and healthcare professionals to be vigilant and seek prompt medical assistance if they experienced typical symptoms of cerebral venous sinus thrombosis (CVST), a potentially fatal clot in the brain [2,3]. European Medicines Agency has also revised the summary of product characteristics and listed thrombocytopenia (very low platelets) as a ‘common’ side effect (i.e., 1 in 100 to 1 in 10) of Vaxzevria, i.e., the CoViD vaccine AstraZeneca [4]. The pharmacovigilance data also suggests that thrombocytopenia is also a frequent observation followed by mRNA CoViD vaccines such as Pfizer or Moderna.
The ‘heavy menstrual bleeding’ has been previously reported in females with underlying platelets disorders [5]. It is plausible that the vaccine-induced thrombocytopenia may be an explanation for the recent incidences of heavy menstrual bleeding experienced by women in different countries after the CoViD-19 vaccination. The significant loss of blood in many women may lead to severe anaemia, further exacerbate thrombocytopenia, and therefore may significantly increase the risk of haemorrhages and clots.
Clinicians and front-line healthcare workers are advised to encourage women to report heavy menstrual bleeding or other extraordinary bleeding events post-vaccination formally into the vaccine adverse events reporting system and seek prompt medical advice. Public health agencies and regulatory authorities are also requested to investigate these incidences and issue further warnings, as this can, possibly be an early sign of potentially fatal vaccine-induced prothrombotic thrombocytopenia leading to rare CVST events in younger women. There may be some women with pre-existing conditions or those on certain medications that may be at increased risk of experiencing post-vaccination severe adverse events and early warnings will help saving lives.
References:
[1] https://www.gov.uk/government/publications/coronavirus-covid-19-vaccine-...
[2] https://doi.org/10.1186/s40545-021-00315-w
[3] https://www.ema.europa.eu/en/news/covid-19-vaccine-astrazeneca-benefits-...
[4] https://www.ema.europa.eu/en/documents/product-information/vaxzevria-pre...
[5] https://www.sciencedirect.com/science/article/pii/S1083318816001637
Competing interests: No competing interests
Dear Editor
The recent reports of cerebral venous sinus thrombosis (CVST) following administration of CoViD-19 viral vector vaccines (AZ/Oxford and J&J/Janssen) have a peculiar clinical presentation exhibiting haemorrhage, blood clots and thrombocytopenia.
We previously proposed a mechanism [1-2] to explain the vaccine-induced prothrombotic immune thrombocytopenia (VIPIT) and reported that the genetic CoViD-19 vaccines (both viral and non-viral vector-based) may directly infect platelets or megakaryocytes triggering mRNA translation and consequent spike protein synthesis intracellularly. This may potentially result in an autoimmune response against platelets and megakaryocytes. The consequent thrombocytopaenia may lead to internal bleeding and spontaneous blood clots. We also proposed that the increased circulatory levels of acute-phase proteins, as observed in the pre-clinical vaccine studies in animals, may also be a contributory factor in putting the haemostatic system at an increased thrombotic potential [3].
The pharmacovigilance data confirmed the CVST incidences with all genetic vaccines (viral or non-viral vector), however, the regulatory authorities in their recent investigations reported that the CVST was unusually accompanied with thrombocytopenia in subjects injected with CoViD-19 viral vector vaccines (such as AstraZeneca and J&J/Janssen) than those injected with mRNA vaccines. We, therefore, looked at the preclinical studies of these vaccines to ascertain their biodistribution to body tissues (for instance brain) beyond the injection site for a possible explanation of the rare fatal clots formed in the brain.
Although, the modern viral vectors that are used in CoViD vaccines are silenced (replication-deficient), each dose of the vaccine contains a very high viral load (e.g., 50 billion viral particles per dose in Ox/AZ or J&J/Janssen CoViD-19 vaccines whereas 100 billion viral particles per dose in the Sputnik-V). The viral particles are unlikely to be confined to the muscles at the injection site; they are free to distribute across the body and drain through the lymphatic system; their apparent volume of distribution is likely to be very high. The biodistribution of ChaAdOx1 containing HBV in BALB/c mice (study 0841MV38.001) indicated the highest viral levels at the injection site, but low levels of virus were still detected after 24 hours of injection in all other tissues (including blood, brain, heart, inguinal lymph node, kidney, liver, lung, gonads, and spleen). The proportional tissue distribution of viral vectors in the body tissues away from the injection site was likely to increase with time, however, biodistribution beyond 24h post-dose was not studied. The biodistribution of ChAdOx1 encoding nCoV-19 following intramuscular injection in mice (study 514559) was ongoing at the time of its regulatory approval [4]. The study 514559 was aimed to examine the biodistribution of ChAdOx1 nCoV-19 in bone marrow, brain, spinal cord, sciatic nerve, and other body tissues. The data from this study is not yet available in the public domain but this might provide evidence of vaccine delivery in the brain. We, therefore, agree with your comments that all vaccine-related data and analyses in possession of the regulatory authorities must be published in full without any further delays.
However, in the absence of the results of study 514559, the biodistribution of ChaAdOx1 HBV in mice (study 0841MV38.001) confirms the delivery of vaccine into the brain tissues. The vaccine may therefore spur the brain cells to produce CoViD spike proteins that may lead to an immune response against brain cells, or it may spark a spike protein-induced thrombosis. This may explain the peculiar incidences of the fatal CVST observed with viral vector-based CoViD-19 vaccines. There is very little information in the public domain to assess the biodistribution of all genetic vaccines, however, it is anticipated that if it is characteristic to the viral vector employed in the vaccine, then the other vaccines using similar technology may also lead to the same safety concerns. Some examples of these vaccines include AstraZeneca/Oxford (Chimp adenoviral vector), J&J/Janssen (Human adenoviral vector 26), CanSinoBio (Human adenoviral vector 5), and Sputnik V (Human adenoviral vectors 26 and 5).
For COVID-19 mRNA Vaccine (Pfizer or Moderna), the biodistribution studies in animals were not conducted. The surrogate studies with luciferase and solid-lipid nanoparticles (Pfizer) confirm a biodistribution to the liver and other body tissues beyond the administration site [5]. For Moderna, the biodistribution of mRNA-1647 (encoding CMV genes) formulated in a similar lipid nanoparticulate delivery system confirms a biodistribution beyond the injection site, in particular, the distribution to the lymph nodes, spleen and the eye was noted [6]. However, the detailed tissue-specific distribution of mRNA vaccines encoding SARS-CoV-2 spike proteins (Pfizer or Moderna) is not fully known that can offer invaluable insights into the potential safety of these vaccines in peoples with pre-existing conditions or those on certain medications.
The detailed biodistribution data including pharmacokinetics of various CoViD vaccines were not conducted by the vaccine manufacturers because the studies demonstrating biodistribution of antigens were considered ‘not required' by the regulatory authorities on the premise that vaccines work by an immunological response than the classic pharmacological approach. However, such an exemption may barely justify the conventional vaccines such as those incorporating whole inactivated virus, split virion, or the sub-unit vaccines, that directly attracts an immune response post-injection.
On the contrary, modern genetic vaccines work on the premise of gene delivery, therefore, a detailed biodistribution and pharmacokinetic evaluation of the formulated product is invaluable in understanding the potential impact of vaccine encoding gene transfection to various body tissues beyond the site of injection. Vaccines are one of the great discoveries in medicine that has improved life expectancy dramatically. However, if genetic vaccines were to be sustained beyond the CoViD19 pandemic, a tissue targeted approach may be the way forward to limit the antigen (the encoding gene) distribution to the intended tissues only to improve the vaccine safety profile for a global mass public rollout. In comparison, the conventional vaccine approaches (classic non-genetic formulations) have a long history of human use across much wider age groups (infants to elderly) and have an established safety profile despite the current challenges in antigen propagation and large-scale production in a timely manner using conventional methods.
References:
[1] https://www.bmj.com/content/372/bmj.n699/rr-6
[2] https://www.bmj.com/content/372/bmj.n699/rr-20
[4] https://www.ema.europa.eu/en/documents/assessment-report/covid-19-vaccin...
[5] https://assets.publishing.service.gov.uk/government/uploads/system/uploa...
[6] https://assets.publishing.service.gov.uk/government/uploads/system/uploa...
Competing interests: No competing interests
Dear Editor
We extend our sympathies go to the patients (and their families) who sadly died after COVID-19 vaccination due to blood clots (now reported both for Astra Zeneca and J&J vaccines). We are glad to see the pauses in administration of these vaccines and support common sense recommendations to prevent further deaths. Those who died have been mostly young women, likely with no pre-existing medical conditions. What would have been their probability of dying of COVID-19? Close to zero.
Prof. Hunter argues in his opinion for a “full speed ahead” approach to vaccination. He states, “For comparison, in women taking hormonal contraceptives the risk of thrombosis is about 60/100,000 person years and risk of fatal pulmonary embolism is about 1/100,000.” This is true but differs from our current dilemma. Those women are making individual choices based upon informed consent. Sadly, the vaccination programs do not to date involve similar transparency and informed consent. The goal of any intervention – including vaccination – is Primum non nocere "first, do no harm."
The discovery of vaccination-related clotting, though very rare, has to become an important part of the conversation. There are reasons for young persons to continue to use these vaccines – a young adult with COVID-19 is extremely unlikely to die of her disease but can transmit disease to a more vulnerable person whose risk of death is far higher. How do we balance the death of one young person against the potential saving of other lives?
Vaccination must continue aggressively for those whose risk of dying from COVID-19 are high, arguably with any of the available vaccines; with those vaccinated fully informed of every known risk.
Competing interests: No competing interests
Dear Editor
Good to see this balanced piece from Prof Hunter and good to see the Netherlands study referred to. This is probably the most appropriate data on pre-Covid CVST incidence available and suggests the 22 cases found in the first 15.8 million first doses of AZ vaccine is significantly lower than expected in an unvaccinated population of that size. I provided further information on this in my Response to a recent BMJ editorial including the reference to the original study.[1]
1. Mahase E. AstraZeneca vaccine: Blood clots are “extremely rare” and benefits outweigh risks, regulators conclude. BMJ 2021; 373 doi: https://doi.org/10.1136/bmj.n931 (Published 08 April 2021) Cite this as: BMJ 2021;373:n931
Competing interests: No competing interests
Re: Thrombosis after covid-19 vaccination
Dear Editor
Having read the responses, specially from Camilo and Merchant, I am constrained to reserve my judgement about the comparative safety of Astra Zeneca, as against Pfizer.
Competing interests: Have had two AstraZeneca jabs